MOCCHA Analysis of Dynamic, Cloud, and Aerosol Processes

Clouds are the dominant control of the surface energy budget through their impact on both solar and infra red radiation. The effect of clouds on the surface radiation budget is commonly expressed as the ‘cloud radiative effect’ (CRE): the difference between the radiation under clear skies and with clouds present. Results from the latest Climate Model Intercomparison Project (CMIP5) show that the annual mean CRE of all the models is biased by -10 W m-2 with a standard deviation across the models of similar magnitude. The bias for the summer months only is +20 W m-2. A bias of just 1 W m-2 over a year is sufficient to melt/freeze a 10 cm thick layer of sea ice; these model biases are thus highly significant. We aim to reduce the bias and uncertainty resulting from model summertime clouds in the Arctic, by improving understanding and parameterization of key processes affecting their formation, development, and dissipation. Arctic summer boundary-layer clouds are mixed-phase – containing both liquid water and ice crystals. This is a thermodynamically unstable non-equilibrium state because saturation water vapour pressure over ice is lower than that over water, thus ice crystals should grow at the expense of water droplets. The clouds are maintained in a quasi-equilbrium through a complex set of interactions between aerosol (on which droplets and ice crystals form), thermodynamic, microphysical, radiative, and turbulent mixing processes. All of these individual processes have significant uncertainties associate with them, and current model parameterizations have been developed almost entirely based on measurements in mid-latitude clouds where conditions are very different. We will use a combination of state-of-the-art measurements of cloud and boundary layer properties and processes, with detailed modelling studies using the Met Office Unified Model. The measurements utilise high resolution active remote sensing systems such as cloud radar and lidar, processed through the CLOUDNET algorithm to generate detail cloud property data. These are supplemented with in situ measurements of boundary layer structure and aerosol properties. Analysis of combined model and observational data will be used to identify key failings in model parameterizations, and new understanding of the physics that must be represented in the models. An iterative process of parameterization updates and testing against observations will be used to improve the model physics. The updates will then be tested in 20-year climate model runs using cases from the Atmospheric Model Intercomparison Project (AMIP) to evaluate their impact on climate projections. The Unified Model is chosen as an ideal tool because of its seamless philosophy, using the same model physics in both operational forecast and climate models. This allows direct testing against observational cases, and easy transfer of new model updates to the climate version. The final model updates will then be put forward to the Met Office for integration into the official model releases.